Skill ScannerOpen-source security scanner that audits AI agent skills for prompt injection and malicious patterns.

4.7 (6)

Прегледано от Daniel Nikulshyn·Актуализирано май 2026 г.

Open Source Security Static Analysis AI Agents DevSecOps CI/CD Developer Tools

Преглед

Skill Scanner is an open-source static analysis tool built to inspect AI agent skills and plugins for security risks before they are deployed. It scans skill manifests, instructions, and bundled code for signs of prompt injection, hidden data exfiltration attempts, and suspicious code patterns that could compromise an agent or its users. Results are emitted in SARIF format, making it straightforward to integrate findings into CI pipelines, code review workflows, or security dashboards like GitHub code scanning. Developers and security teams can use it to vet third-party skills, harden their own, and enforce baseline checks across an agent ecosystem. Because the project is open source, rules and detectors can be extended or customized to fit organization-specific threat models and policies.

Ключови функции

Prompt injection pattern detection
Data exfiltration heuristics
Malicious code pattern scanning
SARIF report output
CI/CD pipeline integration
Extensible rule set

Цени

Модел: Freemium
Категория: Software Testing (QA) Agents
Оценка: 4.7 / 5 (6)

Случаи на употреба

Vet third-party agent skills before deployment

Scan external skills and plugins for prompt injection patterns and suspicious code before adding them to your AI agent, reducing the risk of compromised integrations.

Enforce security gates in CI/CD

Integrate Skill Scanner into build pipelines using its SARIF output to automatically block pull requests that introduce risky skill manifests or instructions.

Surface findings in GitHub code scanning

Pipe SARIF reports into GitHub code scanning or security dashboards so developers and security teams can triage agent skill vulnerabilities alongside other code issues.

Harden in-house skills with custom rules

Extend the rule set to match organization-specific threat models, ensuring internally built agent skills meet baseline checks for data exfiltration and malicious patterns.

Плюсове и минуси

Плюсове

Free and open source
Targets agent-specific threats like prompt injection
SARIF output integrates with existing security tools
Useful for CI/CD security gates
Customizable detection rules

Минуси

Requires technical setup and CLI familiarity
Static analysis cannot catch all runtime attacks
Coverage depends on community-maintained rules

Отзиви

4.7

Средно от 6 оценки.

Влез, за да оставиш отзив.

Fatima Zahra

Mar 22, 2026

Does the job

Pretty happy overall. Extensible rule set just works and sARIF output integrates with existing security tools. Static analysis cannot catch all runtime attacks can be annoying, but no dealbreakers — I'd recommend it to a friend without hesitating.

Carlos Mendoza

Mar 5, 2026

Use it every day

Honestly didn't expect to like it this much. SARIF report output is exactly what I needed, and free and open source. I do wish static analysis cannot catch all runtime attacks, but I reach for it almost every day now and it just clicks.

Mei-Ling Wong

Jan 26, 2026

Use it every day

Honestly didn't expect to like it this much. SARIF report output is exactly what I needed, and useful for CI/CD security gates. but I reach for it almost every day now and it just clicks.

Robert Ainsworth

Jan 16, 2026

Years in this space

I've evaluated a lot of these over the years. What stands out here is sARIF report output — handled better than most — and free and open source. Worth the time if this is your use case.

Yuki Mori

Aug 16, 2025

Use it every day

Honestly didn't expect to like it this much. SARIF report output is exactly what I needed, and targets agent-specific threats like prompt injection. I do wish coverage depends on community-maintained rules, but I reach for it almost every day now and it just clicks.

Kwame Mensah

Jul 13, 2025

Solid for our team

We rolled this out across the team last quarter and useful for CI/CD security gates. Data exfiltration heuristics fits neatly into how we already work, and sARIF report output removed a step we used to do by hand. Requires technical setup and CLI familiarity, which is the main caveat, but it has held up under daily use.

Въпроси

How much does Skill Scanner cost and what's the licensing model?

Skill Scanner is free and open source, so there are no licensing fees. You can self-host and run it as part of your own workflows, with the trade-off that you handle setup, maintenance, and any rule customization yourself.

What threats can it detect, and what are its limitations?

It performs static analysis on skill manifests, instructions, and bundled code to flag prompt injection patterns, data exfiltration heuristics, and suspicious code. As a static tool, it can't catch all runtime attacks, and detection quality depends on the community-maintained or custom rule set.

How does Skill Scanner integrate with CI/CD and existing security tooling?

It outputs findings in SARIF, the standard format consumed by tools like GitHub code scanning, security dashboards, and code review workflows. This makes it straightforward to wire into CI/CD pipelines as a security gate alongside other static analysis tools.

Задай въпрос

Алтернативи на Software Testing (QA) Agents

CarbonCopies AI

Software Testing (QA) Agents

AI двойници имитират взаимодействия на потребителите, за да извършват автоматизирано тестирование на UX и функционална ефективност и да детектират грешки в програми и уебсайтове.

4.8 (4)

Freemium

Diffblue Cover

Software Testing (QA) Agents

Автономен агент на AI, предназначен за генериране и поддържане на единични изпитания за Java на мащаби с гарантирана точност.

4.7 (6)

Paid

PentAGI

Software Testing (QA) Agents

Open-source autonomous penetration testing agents that run 20+ security tools in an isolated Docker sandbox with memory and web intelligence.

4.6 (5)

Free

Flowtest AI

Software Testing (QA) Agents

Агент за интелектуална помощ, който наблюдава уебсайтове чрез изкуствено симулирано потребителско взаимодействие с цел детектиране на проблеми и гарантиране на работоспособност.

4.4 (5)

Freemium